Fix contextual placement quality for v1.0.1 - #2
Merged
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Improves contextual link-placement quality based on real-world v1.0.0 testing.
This patch fixes whole-word anchor matching, preserves publisher capitalization, handles grammatical anchor variants, and improves destination-intent relevance.
Evidence / behavior
Real-world testing with:
Source:
https://titancodes.com/blog/ai-automation-for-small-businesses/
Target:
https://digitalexclude.com/ai/ai-agent-cost/
Requested anchor:
AI Agent
exposed a case where the previous matcher could generate [AI Agent]s when the source text contained “AI agents”.
This update:
Prevents partial-word anchor insertion such as [AI Agent]s
Preserves existing capitalization such as AI agent
Supports natural grammatical variants such as AI agents
Separates Requested anchor from Placed anchor in CLI output
Adds clearer destination-fit evidence
Improves weighting for destination intent such as cost/pricing context
Keeps evidence-first recommendations rather than relying on a universal backlink score
Testing
Tests added or updated where appropriate
32 offline unit tests passed
Python compilation passed
Package build passed
No live-site dependency added to deterministic unit tests
Safety and methodology
No URL/network safety protections weakened
Recommendations continue to expose supporting evidence
New destination-fit behavior is documented and limited
No secrets or confidential datasets included
Version
Updates Backlink Intelligence from v1.0.0 to v1.0.1.